Contamination on a window of an optical imaging system housing can degrade the image quality obtained by the imaging system. The elimination of this “noise” greatly improves the fidelity of the data being generated. The elimination of optical noise makes the data being generated from an analytics system more reliable and the overall performance of the system better, thereby making it more valuable.
Dirty windows on an imaging system are not desired, and an effective way to detect this automatically is highly desired. A human in-the-loop could conceivably look at each image and make a call as to whether or not the windows needed to be cleaned. This is somewhat impractical for a large installation of thousands of systems. A system operator could just periodically clean all of the windows on a scheduled basis.